Expand my Community achievements bar.

Guidelines for the Responsible Use of Generative AI in the Experience Cloud Community.

In the reprocess function, add an option to just process certain items like presets, metadata, etc

Avatar

Level 4

4/30/21

Request for Feature Enhancement (RFE) Summary: In the reprocess function, add an option to just process certain items like presets, metadata, etc
Use-case: Select an asset or folder and then hit reprocess. In the reprocessing dialog, the only option is "Full process". It will be good if there is an option to reprocess --- metadata, presets, smart crops
Current/Experienced Behavior: Select an asset or folder and then hit reprocess. In the reprocessing dialog, the only option is "Full process".
Improved/Expected Behavior: Have an option to select --- metadata, presets, smart crops
Environment Details (AEM version/service pack, any other specifics if applicable): AEMaCS
Customer-name/Organization name:  
Screenshot (if applicable): image.PNG
Code package (if applicable):  
10 Comments

Avatar

Level 4

5/3/21

There are instances wherein we are just adding smart crops and we want that to be available across all assets. If we do that, we need to re-process everything. The problem with that though is we can't easily do that since it will extract embedded metadata again from the assets and override the values we added like title and description and it will also re-run smart tagging which means, the smart tags we have removed may potentially come back

Avatar

Employee

5/9/21

@mikeetiuPH 

The smart crops can be made available to other assets via the image profiles so not sure how reprocessing of asset can be used here to make the smart crop available cross all assets (?)

Status changed to: Needs Info

Avatar

Level 4

5/9/21

What is the purpose of the dropdown in the dialog if the only option is to run a full reprocessing? 

@Jörg_Hoh  - It is important for us because we are regularly adding new presets and smart crops. How are we going to have those new presets and smart crops available in our new assets if we cannot use the "reprocessing" action of assets since it will affect the metadata in our assets? 

Avatar

Employee Advisor

5/14/21

so you mean that the reprocessing is overwriting metadata properties, which you have customized already?

 

In that case turn on the xmp writeback, which will modify the binaries and update them with your adjusted values; then the reprocessing won't overwrite your customized values anymore.

Avatar

Level 4

5/14/21

Can we get an official confirmation that the XMP Writeback is going to be a permanent thing in AEM Cloud? Note that we have asked our Adobe contacts and they are not yet sure if XMP write-back is a good option or permanent solution. XMP Writeback as I understand was disabled by default because it causes performance implications. Thanks so much! 

@jsalanga 

Avatar

Level 1

5/14/21

@Jörg_Hoh - from the AEM Cloud documentation, it is currently not supported:

https://experienceleague.adobe.com/docs/experience-manager-cloud-service/assets/assets-cloud-changes...

 

I know in the past this is something that must be turned on manually due to performance issues. Has this changed with Cloud and is it supported?

Avatar

Employee

5/14/21

@mikeetiuPH @jsalanga 

If you have a custom processing profile assigned to a folder, the reprocess popup shows the option to just reprocess based on that PP

image.png

The docs weren't updated after we added the writeback feature, but it is supported in Cloud Service. If you don't care to writeback metadata, my recommendation would be to actually create custom properties for your use case to prevent the overwriting. For example, if asset1.jpg already has a dc:title XMP property in the binary, AEM will extract it, then you overwrite it in AEM to "my custom title", but don't writeback to the xmp, next time a metadata extraction happens, dc:title will be overwritten. You should add a custom property to the schema "myTitle" and update that property. 

Writeback in Cloud Service is handled by Asset Compute Service, so the scalability/performance issues are not there, but it is asynchronous and we have a limit of how big the metadata content body can be. 

/cc @Jörg_Hoh